Manual swapping a Regina car

[DaleGribble]

Hello everyone, I just managed to manual swap a 1992 Volvo 940 GL Regina car. I've manual swapped a dozen Redblock volvos at least, but I went into this swap with the owner telling me it's a 3.1 vehicle. I sourced everything for the swap and had everything going flawlessly up until I mounted the LH2.4 flywheel. The car would backfire and fire on BDC so I immediately knew something was off. Once I counted the windows and blanks on the oem flexplate and the 2.4 flywheel I realized I was in for a treat. I struggled for about 3 hours trying to find a work around to this issue, until I realized how the Regina CPS worked. I can't speak for any bosch LH but for the Regina cars CPS, I realized it didn't give a shit about the number of windows, it only cares about 2 sets of blanks 180° offset from one another. After realizing this, I filled 2 of the windows opposite the blanks on the 2.4 flywheel with some .030 mig. Low and behold the Regina car loved it. I can now say I've manual swapped a Regina 940 successfully without any modifications outside of filling 2 blanks on a 2.4 flywheel. I hope this information helps anyone who wishes to have a manual swapped Regina car, my friend Is loving his m47 swapped 940 and couldn't be happier with the results. To make filling the blanks easier, the opposing blanks will be 28 windows away on either side of the oem blanks on the 2.4 flywheel. Pm me for a picture, might help explain it better than I can with words.

 

[DaleGribble]

Example A in the picture is a Regina style flex plate orientation, example B is for a standard 2.4 flexplate/flywheel. Filling the windows with Mig welds is the easiest solution I could come up with and hopefully anyone can benefit from this information.

 

[ortho stice]

I remember someone had a Regina flywheel for sale in the classifieds for a long time but this is a solution I’d never heard of!

 

[esmth]

Have you put a timing light to this?

The way these trigger wheels tend to work is after a synchronization period, (the verification that the time from first tooth after a missing tooth gap is twice the previous tooth) it simply counts teeth and determines where inside that tooth count after the gap the coil should be charged and fired. I can't math it in my head without coffee but ASSuming you installed the missing tooth gap in the same spot as it should be with the regina flywheel, your timing is going to be 36% too advanced. (Someone pls double check this!) The ecu might think the engine is spinning that % faster too. (depending on if it uses time between teeth or time between missing teeth. EZ116K uses 15 small teeth.) So it might be reading in a very wrong place in the mapping.

 

[DaleGribble]

Have you put a timing light to this?

The way these trigger wheels tend to work is after a synchronization period, (the verification that the time from first tooth after a missing tooth gap is twice the previous tooth) it simply counts teeth and determines where inside that tooth count after the gap the coil should be charged and fired. I can't math it in my head without coffee but ASSuming you installed the missing tooth gap in the same spot as it should be with the regina flywheel, your timing is going to be 36% too advanced. (Someone pls double check this!) The ecu might think the engine is spinning that % faster too. (depending on if it uses time between teeth or time between missing teeth. EZ116K uses 15 small teeth.) So it might be reading in a very wrong place in the mapping.

Car has perfect timing and runs better than it did before the swap. Timing light shows standard factory timing. Cps didn't seem to care about number of windows so much as the number and position of "blanks"

 

[DaleGribble]

Car has perfect timing and runs better than it did before the swap. Timing light shows standard factory timing. Cps didn't seem to care about number of windows so much as the number and position of "blanks"

Granted I did retime the motor after the swap, due to the owners improperly done diy timing job

 

[dl242gt]

That's some great diagnosis and solution. Thanks for sharing the info!

 

[esmth]

Car has perfect timing and runs better than it did before the swap. Timing light shows standard factory timing. Cps didn't seem to care about number of windows so much as the number and position of "blanks"

tbh I doubt that is the case. I guess it is possible but extremely doubt that regina ignition is internally using distributor-like 1 reference pulse per ignition event logic. again assuming the missing tooth spot is the same spot in reference to crank, the red arrow is where the ecu thinks top dead center is by counting 12 (44-2 pattern) teeth, but in reality it is 4 60-2 pattern teeth (24 degrees!) too early.

 

[DaleGribble]

tbh I doubt that is the case. I guess it is possible but extremely doubt that regina ignition is internally using distributor-like 1 reference pulse per ignition event logic. again assuming the missing tooth spot is the same spot in reference to crank, the red arrow is where the ecu thinks top dead center is by counting 12 (44-2 pattern) teeth, but in reality it is 4 60-2 pattern teeth (24 degrees!) too early.

I'll send some videos today

 

[esmth]

I am not sure if I am thinking about this correctly but the ignition error should increase the closer to TDC the commanded ignition angle:

ecu commanded ignition angle btdc	real (crank) ignition angle btdc	ignition angle error
90	90	0
60	68	+8
45	57	+12
30	46	+16
15	35	+20
12 (idle)	32.8	+20.8
6	28.4	+22.4
3	26.2	+23.2
0	24	+24

Again assuming tooth gap in same spot.

 

[shoestring]

This is interesting. Reading the system operation description, you would think that the EZ does in fact read pulses (how else would it know that it's 90* after the long tooth). However, is it possible it's counting teeth between the long teeth, and doing the math? I only ask this because the OP has made a car that did not run, run. Cam and I were discussing this car yesterday and neither of us could come up with a solution for this aside from "find the marine flywheel". I've never seen a Regina car let alone know how it might work. If the car now runs, and the timing, at least at idle, is correct, I'd say you're on to something. The danger here is being right for the wrong reason...

 

[esmth]

I doubt it is dynamically counting teeth and using that. that would need a large amount of logic/code space in the interrupt handler I don't think they could afford in the 80s when designed. But I could be wrong... I think more likely if OPs timing was actually correct, it is using the missing tooth as a simple crank reference mark and otherwise operate sort of like the hall sensor style ignition internally. But then why did they bother to drill 40 holes instead of 2 if it doesn't actually count them?

I don't have a regina car or box handy either so I can only guess but I assume what is happening in OP's case is as follows: The ignition box sees the long tooth and synchronizes with the crank. Engine runs 'normally' (with ignitiontiming off of course.) But once the ECU counts to the 22th tooth, it expects a long tooth, which it never gets. It looses crank sync and starts searching again for the long tooth again. It then finds it some more teeth later. Repeat. Engine will still run, just uhh debatably correctly

Here is a logic analyzer trace of me cranking and starting my car with ez116k (60-2) this evening. You can see right after the FIRST missing tooth gap on the top black trace, the EZK starts to fire the coil. I assume the regina ignition will do the exact same after it sees it's first missing tooth, even if it hasn't verified the actual total tooth count. And because there is now an excess of teeth in OP's case, it'll keep firing the coil and then losing sync, re-finding the long tooth & sync, and fire the coil in a loop. With the symptom being too early ignition timing.

 

[DaleGribble]

I doubt it is dynamically counting teeth and using that. that would need a large amount of logic/code space in the interrupt handler I don't think they could afford in the 80s when designed. But I could be wrong... I think more likely if OPs timing was actually correct, it is using the missing tooth as a simple crank reference mark and otherwise operate sort of like the hall sensor style ignition internally. But then why did they bother to drill 40 holes instead of 2 if it doesn't actually count them?

I don't have a regina car or box handy either so I can only guess but I assume what is happening in OP's case is as follows: The ignition box sees the long tooth and synchronizes with the crank. Engine runs 'normally' (with ignitiontiming off of course.) But once the ECU counts to the 22th tooth, it expects a long tooth, which it never gets. It looses crank sync and starts searching again for the long tooth again. It then finds it some more teeth later. Repeat. Engine will still run, just uhh debatably correctly

Here is a logic analyzer trace of me cranking and starting my car with ez116k (60-2) this evening. You can see right after the FIRST missing tooth gap on the top black trace, the EZK starts to fire the coil. I assume the regina ignition will do the exact same after it sees it's first missing tooth, even if it hasn't verified the actual total tooth count. And because there is now an excess of teeth in OP's case, it'll keep firing the coil and then losing sync, re-finding the long tooth & sync, and fire the coil in a loop. With the symptom being too early ignition timing.

This is such a cool picture of the live data. Honestly I have no idea why the cps fell for the old mig gun trick, I'm just happy it did. Car has full powerband, no misfires or stuttering like it did when it only had 2 blanks instead of 4. I'll bust out an oscilloscope at some point and try to post some more useful replies than "I'm happy it worked."

 

[DaleGribble]

I also plan to "properly" resolve the issue. I'll pull the flywheel, slap her on a metal lathe, knock off 1/4" and slide on the tone ring from his oem flexplate. Might be a smart idea for me to get some live data before and after the "fix".

 

[dalek]

Maybe JohnV has some blank flywheels

 

[DaleGribble]

Pictures of the 940 Regina and my 740 turbski m47 in both cars

 

[DaleGribble]

940 hit 70mph on the way here. Runs cherry

 

[bobxyz]

I doubt it is dynamically counting teeth and using that. that would need a large amount of logic/code space in the interrupt handler I don't think they could afford in the 80s when designed. But I could be wrong... I think more likely if OPs timing was actually correct, it is using the missing tooth as a simple crank reference mark and otherwise operate sort of like the hall sensor style ignition internally. But then why did they bother to drill 40 holes instead of 2 if it doesn't actually count them?

It's possible that an early 80s microcontroller could count teeth when running if the MCU includes some usable counter/timer logic (e.g. Motorola 6801 or various 8051-based MCUs). Based on the observed behavior, it's more likely that the timing during running is delayed from the gaps without counting teeth.

During cranking, the RPM varies a huge amount in a single revolution. During this time, the ignition box is most likely counting teeth so that spark during cranking can be fired a few degrees after TDC. Once the engine is running, it would switch to timing just off the gaps.

Here is a logic analyzer trace of me cranking and starting my car with ez116k (60-2) this evening. You can see right after the FIRST missing tooth gap on the top black trace, the EZK starts to fire the coil. I assume the regina ignition will do the exact same after it sees it's first missing tooth, even if it hasn't verified the actual total tooth count. And because there is now an excess of teeth in OP's case, it'll keep firing the coil and then losing sync, re-finding the long tooth & sync, and fire the coil in a loop. With the symptom being too early ignition timing.

Nice data capture! For EZ116K, the falling edge of the RPM strobe occurs at TDC of each cylinder [or ~40us after to be more precise]. Spark occurs on the rising edge of spark out. This capture shows that spark during cranking is firing a little bit after TDC. [One of the bugs in my logger code is that it doesn't measure this correctly. Otherwise, I'd post an example log of cranking-to-run spark timing.]

I also plan to "properly" resolve the issue. I'll pull the flywheel, slap her on a metal lathe, knock off 1/4" and slide on the tone ring from his oem flexplate. Might be a smart idea for me to get some live data before and after the "fix".

I don't know if this will work -- magnetically (for the VR sensor), the flexplate has holes with no metal behind them. If you transfer the tone ring, can it still stick out with no metal behind the holes?

Like the EZK setups, the Rex ignition box sends a tach signal to the fuel ECU and gets a Tq load signal back from the ECU. It would be interesting to see how the Rex load signal behaves and if it's compatible with EZ116K. If it's compatible, t might be possible to run a EZ116K box with a 60-2 flywheel instead of the Rex box. The EZK box might need a special chip if the load mapping isn't exactly the same.

 

[DaleGribble]

Whatever it is the car runs cherry, so I'm gonna let it be. Friend has had no issues and has already put around 100 miles on the swap

 

[DaleGribble]

It's possible that an early 80s microcontroller could count teeth when running if the MCU includes some usable counter/timer logic (e.g. Motorola 6801 or various 8051-based MCUs). Based on the observed behavior, it's more likely that the timing during running is delayed from the gaps without counting teeth.

During cranking, the RPM varies a huge amount in a single revolution. During this time, the ignition box is most likely counting teeth so that spark during cranking can be fired a few degrees after TDC. Once the engine is running, it would switch to timing just off the gaps.



Nice data capture! For EZ116K, the falling edge of the RPM strobe occurs at TDC of each cylinder [or ~40us after to be more precise]. Spark occurs on the rising edge of spark out. This capture shows that spark during cranking is firing a little bit after TDC. [One of the bugs in my logger code is that it doesn't measure this correctly. Otherwise, I'd post an example log of cranking-to-run spark timing.]


I don't know if this will work -- magnetically (for the VR sensor), the flexplate has holes with no metal behind them. If you transfer the tone ring, can it still stick out with no metal behind the holes?

Like the EZK setups, the Rex ignition box sends a tach signal to the fuel ECU and gets a Tq load signal back from the ECU. It would be interesting to see how the Rex load signal behaves and if it's compatible with EZ116K. If it's compatible, t might be possible to run a EZ116K box with a 60-2 flywheel instead of the Rex box. The EZK box might need a special chip if the load mapping isn't exactly the same.

I'll lathe the piss outta the flywheel so the tone ring works worst case. Or I'll have PK Machine make the 2.4 identify as a Regina